Nvidia moves Blackwell AI chip production to Arizona after Trump request

Nvidia CEO Jensen Huang announced that the company’s Blackwell AI chips are now being manufactured in Arizona, marking a significant shift from the company’s previous reliance on Taiwan-based production. The move comes in response to President Trump’s request to bring manufacturing back to the U.S. for national security reasons and represents Nvidia’s broader strategy to position itself as essential to American technology infrastructure amid ongoing export restrictions to China.

What you should know: Nvidia has successfully moved production of its most advanced AI chips to U.S. soil after years of exclusive manufacturing in Taiwan.

• The first Blackwell wafers were produced at a Taiwan Semiconductor Manufacturing Company (TSMC) facility in Phoenix, Arizona, with Blackwell-based systems now being assembled in the U.S.
• Huang revealed that Trump had asked him nine months ago to bring manufacturing back, citing national security and economic benefits.
• Demand remains strong with 6 million Blackwell GPUs shipped in the last four quarters, and Nvidia expects $500 billion in GPU sales between Blackwell and next year’s Rubin chips combined.

Strategic partnerships: Nvidia announced a major telecommunications partnership with Nokia, the Finnish telecommunications equipment maker, taking a $1 billion stake in the company to develop 5G and 6G base station technology.

• The collaboration aims to reduce reliance on foreign telecommunications technology, particularly addressing concerns about China’s Huawei dominance in cellular networks.
• Nokia will use Nvidia’s new ARC product, which combines Grace CPU, Blackwell GPU, and networking components in future base stations.
• Huang emphasized that AI delivered over 6G networks could enable robot operations and more accurate weather forecasting.

The China challenge: Export restrictions have significantly impacted Nvidia’s business, with the company currently “100% out of China” according to Huang.

• The U.S. government’s license requirement for Nvidia’s H20 chip—a version specially designed to comply with export controls—cost the company approximately $10.5 billion in sales over two quarters.
• While the Trump administration agreed to approve H20 licenses with a 15% fee on China sales, Nvidia hasn’t revealed a newer China-compatible chip based on Blackwell GPUs.
• The company continues advocating that Chinese AI developers using U.S. technology serves American interests better than forcing China to develop independent alternatives.

Quantum computing expansion: Nvidia introduced NVQLink, a new system connecting quantum chips to its GPUs, addressing national security concerns about quantum computing leadership.

• The technology enables error correction in quantum computing and allows researchers to orchestrate quantum devices with AI supercomputers.
• Seventeen quantum computing startups will produce hardware compatible with NVQLink.
• The company also announced partnerships with the Department of Energy to build seven new supercomputers.

What they’re saying: Huang positioned the announcements as essential for U.S. technological leadership and national security.

• “The first thing that President Trump asked me for is bring manufacturing back,” Huang said. “Bring manufacturing back because it’s necessary for national security.”
• On telecommunications: “Our fundamental communication fabric is built on foreign technologies. That has to stop, and we have an opportunity to do that.”
• “Nvidia is a proud American company building the U.S. AI infrastructure that will ensure our country leads the world in shaping the future of innovation,” said Kari Briski, Nvidia’s vice president of generative AI software for enterprise.